For finding results from websites, you can use Microsoft Copilot, Perplexity, or Google’s Gemini. These tools combine a language model with a source of facts, like web search results. They search the web and use the AI to summarize the text from those web pages, giving you links to where each part of the summary was found.
It’s a good idea to visit those web pages instead of only reading the summary, because it’s still possible for the AI to make errors in those summaries. These tools “hallucinate” less often than tools like the free version of ChatGPT because of this connection to a source of facts. But they can still make mistakes.
For finding results from academic papers, it’s always best to begin with library databases and Google Scholar. However, there are some tools that combine an AI language model with a database of scholarly papers. Elicit is one example. It searches Semantic Scholar and uses AI to summarize results from the journal articles it finds.
With Elicit, you may find some results that you didn’t find with library databases. That’s because Elicit uses semantic searching rather than keyword searching: it can find papers discussing similar concepts even if matching keywords don’t appear. But remember that library databases and Google Scholar index many more journals, so it’s best to use multiple tools if you want to be comprehensive.